von Storch at Boulder

Here is a guest report from Scott Shipley of George Mason University on the von Storch seminar at Boulder.

Hans has published his presentation at the following address:

Click to access 050708.boulder.pdf


I had the pleasure of listening to Hans von Storch yesterday at NCAR. A seminar was quickly called at 3pm, and the Table Mesa conference room was packed (standing room only). Commenting on the addition of a panel of experts to comment after his talk, Hans mentioned that he was uncomfortable with the new approach but it would be ok, since Fred Singer had also been subjected to such treatment (laughter). Hans also observed that the last time he was at NCAR, there were only ten people in the audience, suggesting that this was because of the political sensitivity the topic has acquired.

Hans proceeded to give an enjoyable lecture in two parts, the first examining the failure of the Mann et al. technique but not of the “hockeystick” in general, and the second part examining the politics and societal reaction to the “hockeystick” as a “condensational symbol” (a symbol with emotive power). He stated that we should not be too quick to embrace complex statistical techniques which offer “magic bullets”. He agreed that the McIntyre et al. analysis with “red noise” demonstrates the failure of the Mann et al. technique, but not necessarily of their conclusion. He also counseled everyone to allow time for the scientific method to settle controversies, which may take many more years. His own work results in a “hockeystick”, but with a clear minimum near the Little Ice Age and some “kinks” in the handle. Offered as support for the adequacy of the scientific method, Hans mentioned Spencer & Christy’s result on atmospheric temperature, how the disagreement with surface temperature trends caused the NAS to convene a panel of experts, and how that panel was able to demonstrate a reconcilation of the surface and satellite records. This reconciliation involves some unspecified correction to the Spencer and Christy procedure, of which I am unaware, but which the NCAR audience seemed to take as fact.

The second part of his lecture addressed the difficulties of fighting a condensational symbol. He stated that the Hockeystick has become such a symbol, and similar to the Swift Boat attacks on John Kerry, no amount of patient explanation or truth telling will be sufficient to undo the damage. Citing his own statistics on public opinion (which might have been a sample of as few as seven people?), he stated that public opinion is currently going against climate science. What makes it harder, is that those groups who are arguing for the “hockeystick” (climate scientists, insurance companies, greens) are also the same groups who will benefit the most from acceptance. This argument has been used by the “evil” (sic.) climate sceptics to increase public and government suspicions of the climate majority’s motives.

At the table were Caspar Amman, Doug Nychka, Roger Pielke Jr., and Warren Washington. Doug was introduced as a new mathematics/statistics expert hired by NCAR. He asked if anyone in the audience was uncomfortable with Hans showing only one or two “hockeysticks”, when there should in fact be an ensemble of temperature curves given the uncertainties in the inputs. Few were uncomfortable. When asked about this, Hans explained that he only had two curves, and they were both “hockeysticks”, but that was not the point of his lecture. Doug agreed that Hans’ small sample was sufficient to test his hypothesis, that the Mann technique failed but the “hockeystick” is not ruled out. Caspar went into great detail about how all of his GCM model runs yield hockeysticks. Roger talked about condensational symbols as an academic pursuit. Warren said that in his 42 years at NCAR, he has at least learned that “time will tell”. Warren added light-heartedly that Hans advises us to not believe complex methods bearing magic bullets, but that Hans’ own procedure is quite complex and so should also be discounted? (cordial laughter)


  1. Roger Pielke, Jr.
    Posted Jul 11, 2005 at 6:37 AM | Permalink

    Scott- Some corrections:

    1. The notion of “condensational symbol” and Swift Boat analogy were my comments.
    2. Hans did not mention anything about public opinion, he did say his view was a personal one, not backed by rigorous data.
    3. The last comment in your summary about complex methods was made by Tom Wigley not Warren Washington.
    4. Doug Nychka has been at NCAR for some time.

    Thanks for posting this. We expect also to have a summary up on our site soon.

  2. Michael Jankowski
    Posted Jul 11, 2005 at 7:55 AM | Permalink

    What a shocker there – the GCMs so heavily based on CO2 forcing produce hockey sticks in response to historical CO2 levels.

    What would be useful to me is a discussion as to how the limited amount of proxies used by Mann, even if analyzed scientifically and properly, can be used to truly represent a global picture.

    As a preliminary analysis, has anyone more familiar with the exact proxy locations examined the 20th century data of the weather stations closest to those proxies to see how well they would represent the global temps of the 20th century?

  3. Tilly Morgan
    Posted Jul 11, 2005 at 8:35 AM | Permalink

    This has nothing to do with the immediate question. But where do you guys stand on ‘peak oil?’
    Is it real? is it a dream? How does it tie in with the questions you are dealing with?

  4. Paul
    Posted Jul 11, 2005 at 12:56 PM | Permalink

    I would like to hear people’s opinion as to why there has been very little (or even no) discussion as to why MBH’s so-called temperature reconstruction was given any credence in the first place, even before M&M’s excellent work. The very act of splicing together various types of temperature proxies and claiming that the result represents past global temperatures in any real sense is absurd on its face. I’m not saying that such an exercise is without scientific merit. Outside the very politicized environment under which climate science now operates, I’m certain that there would have been a small subset of scientists who found the research valuable in a very narrow sense. But the fact that we are now embroiled in a controversy of global proportions over what seems to me a trivial bit of research just shows the complete derailment of climate science into pure politics. The “hockey stick” offered a nice graphic that those scientists with a political agenda knew would look impressive to the lay public, even though they knew that scientifically it signified little, and so they shamelessly promoted it. Every time one of these scientists were called to testify before Congress, the hockey stick was there, blown up to cartoonish proportions. It’s primary purpose was to invoke gasps of horror and long self-righteous speeches from climate catastrophe-believing politicians and to intimidate the doubters into silence. Okay, so I’ve answered my own question, but I’d be interested in other people’s opinions on this.

  5. Doug L
    Posted Jul 11, 2005 at 3:14 PM | Permalink

    I’m not a climate scientist, or a scientist at all, but I think it’s likely that oil production will plateau rather than peak, and that there’s lots of coal to burn to create greenhouse gases.

  6. fFreddy
    Posted Jul 11, 2005 at 3:28 PM | Permalink

    This site is oriented to a very professional analysis of certain aspects of climatology, its consequences for public policy, and certain related issues. I am not aware of our hosts having expressed any opinion on peak oil.

    My purely personal opinion ? Probably tosh. The doom-mongers have been predicting that one thing or another will run out for decades, and they haven’t been right yet. The market adjusts, and new sources are found. The Russians all think East Siberia has even more oil than West Siberia; everyone in the North Sea, which is running out, is convinced there is even more to be found west of Shetland … no doubt when the time comes, they will be given the budget to go find it, and to work out how to get it to market.
    Of course, it would be preferable if some clever person can figure out a better way to store energy than mucky old hydrocarbons, but that’s still another issue.

  7. fFreddy
    Posted Jul 11, 2005 at 5:52 PM | Permalink

    Paul, some rather undisciplined thoughts, which may be rubbish :

    You refer to there being a very politicised environment now. I would suspect that climatology has been very politicised for a decade or two, and that the only difference now is that the subject has become a bit more bipartisan. Previously, I suspect it has been strongly weighted to the left of centre.

    The great contribution made by our hosts here has been in mathematics. My memory of being a maths undergraduate many years ago was that most Natural Scientists didn’t, ummm, display much interest in maths beyond that which was part of their coursework. The reason most climatologists didn’t check MBH ? Most of them didn’t really understand it.

    Climatology is a relatively young area of study. Indeed, arguably, it is not really a separate area even now. Rather it is something that touches on a whole load of more recognised disciplines – particularly meteorology, but also geology, biology, etc. I get the feeling that the average age of practitioners is relatively young – there aren’t really any “grand old names” who might feel some responsibility for maintaining standards.

    Another question to add to this : what is to be done ? If this really is the massive failure of scientific ethics that it appears to be – how can the problem best be remedied, with minimum damage to real science and scientists ?

    Incidentally, isn’t today the dealine for MBH and co to submit data to Congress ? Does anyone know what they’ve done ?

  8. Michael Mayson
    Posted Jul 11, 2005 at 11:06 PM | Permalink

    Off topic – but there is a new blog at http://ccc.atmos.colostate.edu/blog/
    for Roger A. Pielke Sr. Research Group
    From a quick look it has made a good start. Perhaps you could add it to “weblogs and resources” Steve.

  9. Michael Jankowski
    Posted Jul 12, 2005 at 1:28 PM | Permalink


    That’s a point (Mann’s so-called reconstruction) I was trying to get to on the 2nd post of this thread. When I saw a map showing the proxy locations used by Mann, I was amazed that the scientific world would accept that limited coverage as capable of approximating average global conditions. I can’t locate an image showing the proxy locations for Mann’s “reconstructions” of the last 600-1000 yrs, but look at this one for the last 2,000 yrs http://www.ncdc.noaa.gov/paleo/pubs/mann2003b/mann2003b.html . Note that with an extremely limited spatial coverage, Mann and Jones still concluded that, “These reconstructions indicate that late 20th century warmth is unprecedented for at least roughly the past two millennia for the Northern Hemisphere.”

    I did question why Mann’s work was even considered a “reconstruction” on realclimate.org a few weeks ago. I was pointed to one of the definitions of reconstruction that said (almost direct quote, if not direct) “an interpretation based on evidence.” I tried a number of times to question how an “interpretation” was considered “science,” but that post never made it through the RC censors.

    I can see how MBH was publishable – provided it were done properly and didn’t bury particular statistics (in fact, it should’ve emphasized them as a caveat). But regardless of whether or not it MBH is significantly flawed, I can’t see how it can be considered to be so conclusive with such a limited coverage (what is MBH’s final number – 112 proxy locations?).

  10. Posted Jul 12, 2005 at 5:29 PM | Permalink

    I discussed this very question with a retired Exxon-Mobil senior executive. He confirmed that the Hubbert’s Peak hypothesis applies to easily won, “sweet” oil, that can economically be recovered at prices less than say US$20 per barrel. He also confirmed my view that the industry, for reasons that it best understands, persists in avoiding the proper definitions of “reserves”. In mining “reserves” means that portion of the resource that can be extracted economically with current technology. It follows that the “reserves” of oil are calculated at an implied price, probably US$20 per barrel according to my information. The real question is what are the available reserves of oil at a long term sustainable oil price of US$40 per barrel for example. My colleague confirmed that there are huge resources of hydrocarbons available that can be turned into oil if the price is sufficiently high. For example, there are very large undeveloped reserves in Russia that require very substantial capital investment, something that the oil majors are not prepared to do in the current political/economic environment in Russia. Also, there are further very large reserves in the Athabasca tar sands, and the Venezuelan Orinoco bitumen deposits. There are also opportunities to convert gas to oil, coal (lignite and brown coals) to oil etc. These all require very large capital investments.

    The issue is more a price and capital markets issue than it is a resource constraint. If the financiers could rely on a guaranteed oil price of say US$40 or US$45 per barrel for a 20 year period, then the capital would certainly be available to develop the resources I have mentioned. The difficulty is that it would likely soon result in over supply which would see the market price drop below those prices, placing pressure on the financing. So the real answer to oil supply is to find a way to guarantee oil prices of US$40 or US$45 per barrel for 20 years. You will have all the supply you want if you do. Volatility and price risk are the main factors limiting development of alternative fuels.

  11. Grid
    Posted Jul 13, 2005 at 10:08 AM | Permalink

    though I may agree with some points on here, I do not see the point in spewing out false allegations in the name of rallying. Atmospheric scientists at most – if not all – american universities are required to take 4 terms of calculus studies for undergraduate studies, and likely 2 more as graduate students. You are correct that many environmental scientists are not mathematicians, but “physical” scientists are and they are the ones ‘studying’ the climate.

    Also, there are many “old timers” as you put it in the climatology field. Perhaps you should do some research on that fact, particularly with GFDL, before you claim otherwise. To name some recently heralded older climatologists: Ed Lorenze and Suki Manabi.

    I promote your effort in playing scientists yourselfs, but please refrain from making erroneous remarks.

  12. Michael Jankowski
    Posted Jul 13, 2005 at 11:23 AM | Permalink

    What american university atmospheric science programs are you familiar with? In my experience, most schools only offer at most 4 semesters of calculus for non-math majors (anything beyond that is almost purely theoretical…I don’t quite see how proving mostly abstract mathematical theorems would be of much benefit to climate scientists), with the first 2 semesters being comparable to a quality year of high school calculus.

    Specifically, in regards to MBH, we’re mostly talking about statistical issues. IMHO, statistics is heavily under-represented in education, yet it is vitally important to scientists – particularly in one such as climate with such potentially huge margins of error. The emphasis seems to be usually on probability distrubutions and statistics revolving around the normal distribution. I’ve seen some climate scientists argue statistics with people who have a statistical background – it isn’t pretty. Yet these same climate scientists adamantly insist they are correct.

    FWIW, following is the listing of some of the requirements to be admitted to the graduate program in atmospheric science at Columbia University. James Hanson is listed as a professor, Gavin Schmidt over at realclimate.org is a research scientist, etc, and the association with NASA’s Goddard Institute of Space Studies (GISS) is certainly noteworthy. So CU is far from chump change in the climate science realm.

    “A participant needs a strong background in physics and mathematics, including advanced undergraduate courses in mechanics, electromagnetism, advanced calculus, and differential equations. Deficiencies should be remedied during the first year. Undergraduate courses in atmospheric sciences, earth sciences, and astronomy are helpful but not necessary.”

    Interestingly enough, no mention is made at all of statistics. Yet we are to believe that many scientists are experts in applying statistical methodology to analyze data.

    Also FWIW, Columbia’s only calculus courses are at the undergraduate level and there are only 4 semesters (either calc I-IV, or in combination with linear algebra in honors math I-IV).

  13. fFreddy
    Posted Jul 13, 2005 at 2:03 PM | Permalink


    Grid, I do not understand your comment about “in the name of rallying”. Paul asked a question, I answered it. If you see this as something to do with rallying, then I think you are imposing your own worldview onto what I said.

    “spewing out false allegations” – am I right in thinking an allegation is a statement of fact ? I think you will find that most of my post above was more in the nature of vague thoughts and impressions, which I freely admitted could be wrong. Your clue to this is where the first line says “which may be rubbish”.

    Of course, the reference to my memory of my own university days was a statement of fact. I’m sure you weren’t calling that a false allegation, because that would be calling me a liar, and you are far too polite to do such a thing. I should clarify that when I said “many years ago”, I meant more than twenty years ago. I have no knowledge of whether there was a course in atmospheric sciences back then.

    Regarding old-timers, is that the original chaos theory Lorentz ? If so, what I was suggesting is that he might think of himself as a Grand Old Man of Meteorology rather than a Grand Old Man of Climatology. Please note that this is not a statement of fact, because I don’t know the guy – hence the use of “I get the feeling” in my original post.

    I hadn’t heard of Suki Manabe. A quick Google brings up the quote :
    “I come here from Tokyo and they give me millions and millions of dollars
    of computers so I can enjoy my hobby.”
    Well, I look forward to seeing a precise and concise write-up of what he is modelling with those computers, so I can develop my own opinion of it.

    Your phrase “your effort in playing scientists” is rather patronising. This tends to annoy people, which tends to impede the free flow of discussion.
    Incidentally, the plural of ‘yourself’ is ‘yourselves’, not ‘yourselfs’ as you had it. You might like to broaden your reading so you don’t make such such simple spelling errors.
    See what I mean ?

  14. Scott Shipley
    Posted Jul 14, 2005 at 9:26 AM | Permalink

    Roger: I stand corrected, thanks. -scott

  15. Jeff Norman
    Posted Jul 15, 2005 at 10:31 AM | Permalink

    Michael Mayson,

    I find your NOAA Paleoclimatology link somewhat annoying. Not the link itself but the content at page linked. Some comments/questions.

    1. In “Fig. 2a,b Hemispheric temperature reconstructions” are those blue areas supposed to represent error/uncertainty margins? If yes it appears the error margins are the same in both hemispheres despite the dearth of data points in the Southern Hemisphere(SH). How is this possible?

    2. I notice that the SH’s proxy reconstruction is much more variable than the NH’s and yet it does not show a 20th centrury warming (though this is hard to see, see my point #6).

    3. The NH reconstruction becomes strangely flat btween 300 and 500 AD.

    4. If the SH reconstructed climate between 1850 and 1980(?) does not match the instrument record for the same period does this invalidate the entire reconstruction? In fact it appears as though the reconstruction is not very reponsive to actual measured temperature change. Does this invalidate the methodology?

    5. In the NH there appears to be an inordinant concentration of “PCs of Tree-Ring Networks” (six in fact) in bristle cone country.

    6. “Figure 1” shows a spaghetti diagram. Given the flexibility of the Internet one would think that they could present the graphs individually or grouped together as desired by the viewer with or without the expletive “Instrument record” smudging out the 20th century. Are they trying to hide something? They are succeeding.

    7. Why are the “Mann and Jones with uncertainties” back to 200 AD uncertainties smaller than the “Mann et al. 1999 with uncertainties” back to 1000 AD uncertainties? Are the calculation of these uncertainties described some place?

    8. Doesn’t anyone actually look at this stuff?

  16. Jeff Norman
    Posted Jul 17, 2005 at 5:04 PM | Permalink

    I’ve gone on an uncertainty fishing trip over at realclimate and have a nibble.

    I found this link on quantifying uncertainty interesting:

    9.2. information required
    9.2.1. A complete report of a measurement result should include or refer to documentation containing,
    -a description of the methods used to calculate the measurement result and its uncertainty from the experimental observations and input data
    the values and sources of all corrections and constants used in both the calculation and the uncertainty analysis
    -a list of all the components of uncertainty with full documentation on how each was evaluated


    Steve: which realclimate post?

  17. Jeff Norman
    Posted Jul 18, 2005 at 6:56 AM | Permalink


    In the “The lure of solar forcing” thread. My follow up questions haven’t appeared… yet?


  18. TCO
    Posted Sep 20, 2005 at 11:07 PM | Permalink

    econbrowser has a lot of discussion of peak oil.

    Fred: if you really feel that way, are you going short on the futures market? (and why are you smarter than the market…not saying the market is perfect…but why are you smarter than all those invisible hands motivated by gain to drive the futures price).

    Guaranteeing a high price for oil would be a nightmare. the companies invest based on expectations. They can deal with the risk. That is capitalism. No price controls, boys.

    Yeah, it sorta looks like sweet crude is peaking. Part of the real problem is holdups and regulations inhibiting new refineries (with different technology) and also the low sulfur regs. This has driven the heavy-sweet spread from 5$ to $15.

  19. Neal J. King
    Posted Mar 4, 2008 at 5:46 PM | Permalink

    #12, Michael Jankowski:

    It is quite common for physics undergraduates to take an applied mathematics course, that can be one or two semesters long, and covers complex analysis and partial differential equations.

    In addition, all physics graduate students take a two-semester course in electromagnetic theory, which is mostly a grueling exercise in solving linear partial differential equations.

    Finally, all undergraduate students in physics have to take lab classes, in which basic principles of the statistical analysis of measurements are studied on one’s own time (no formal course, but if you don’t figure it out, you flunk). If they go on to do real work in a lab, they usually get a lot more.

%d bloggers like this: